Home Built Autoguider Project

20/10/04 One of the limiting factors effecting successful Astro-photography is the correct tracking of celestial objects by the telescope mount. Although the Synta EQ6 mount that I use is a very robust unit, it does have some shortcomings in this area, suffering from both Periodic Error and other tracking inaccuracies. There are various methods available for improving it's accuracy but these involve modifying the electronic control circuitry as a minimum plus the installation of new gears and motors for the very best results. Although the electronic modifications are fairly easy and low cost to implement, the additional cost of new mechanical innards can double the cost of the mount and would involve completely stripping down the head unit to it's component parts - a task that I do not really wish to carry out!

I determined that an Autoguider would go some way towards resolving some of these issues and as I already had a CCD camera (web-cam) I realised that I could build my own Autoguider for under 50.00 by using an LX200 Emulator and suitable guiding software. Making the necessary connections to the standard HEQ5/EQ6 handcontroller is pretty straightforward as shown in this article.

29/11/04 I have finally had an opportunity to test my home built Autoguider and I am posting the results ‘warts and all’ here! I am pleased to say that the preliminary results are quite encouraging. The following two photographs of Aldebaran were taken with my Canon EOS 300D at prime focus through my 10" Skywatcher Newtonian Reflector mounted on a Synta EQ6 mount. Auto Guiding was accomplished using an unmodified Philips Toucam Pro II webcam at prime focus through my 80mm Skywatcher short tube Refractor and a home built LX200 emulator controlled by Guide Dog software. Both shots were exactly 7 minutes and 10 seconds long at 100ASA. Aldebaran was chosen not for it's photogenic qualities but because it is away from Polaris, was instantly easy to find and was high enough in the sky to avoid most 'seeing' problems. Of more interest are the much dimmer stars around it!

IMPORTANT UPDATE: (01/07/07) There is now an alternative interface available for autoguiding available from Shoestring Astronomy who also have an excellent manual describing how to make this modification

The autoguided shot is clearly better:-

img_0167.jpg (16846 bytes)



img_0166.jpg (17661 bytes)



The close-up of one of the stars below shows how the rather unpleasant periodic and other errors have been greatly smoothed out and it is interesting to see the graph of the errors matching the visual ‘damage’!

img_0167_part.jpg (8882 bytes)

img_0166_part.jpg (7895 bytes)



In the following graphs, the vertical scale is in arcseconds.

pegraph.jpg (44289 bytes)

The rather horrendous periodic and other errors shown in the graph above matches well with the distortion of the stars. Although, as can plainly be seen in the photographs above, periodic and other tracking errors have by no means been eradicated by this modification, I am pleased with the results and I haven’t started to tweak the response of the unit yet so it may well improve.

As the second part of the experiment, I ran a long guiding session with just my 80mm Refractor mounted on the mount and got the following result:-

pec301104.jpg (58887 bytes)

Does this mean that mount's my Periodic Error is not as high as I thought or that the PE varies with load?

It is important to note that the above PE graphs were produced before I made the upgrade to the SkyScan motors and controllers.

Having now upgraded my mount to the SkyScan system which has GoTo and PEC amongst other features, I have adapted the above Autoguider to work with this new system (see here for wiring). A recent test with all three of my telescopes plus my DSLR and Webcam mounted on the EQ6 resulted in some very good improvements. This shot was a two minute exposure of Vega with PEC enabled but unguided. As can be seen, the smaller stars are nice and round:-

vega_unguided_pec_small.jpg (22299 bytes)

The following shot is a pixel for pixel display of the same test photograph of Vega - it is difficult to fault the tracking!

vega_unguided_pec.jpg (71593 bytes)

I also ran a PE check and obtained the following graph:-

vega_pe_plot.jpg (102349 bytes)

I have now upgraded my imaging camera to the recently announced Starlight Xpress SXVF-H9C with an SXV Autoguiding camera so no longer use the LX200 Emulator. This upgraded version of the original SXV-H9C has a faster USB download time amongst other improvements.

The SXV Autoguider is controlled by the main camera which has a standard ST4 autoguider output interface. This new guider gave a serious amount of grief when I first got it and would not alow the telescope to track correctly. After exhaustive testing, I discovered that the control instructions from the SXV unit were incorrect in that two of the control outputs were swapped round resulting in two direvtional instructions to the mount being 90 degrees out of phase!!

Once I had brought this to the attention of Starlight Xpress along with the input from another (USA) user, the firmware was found to be incorrectly coded and a new set of data was despatched to resolve the problem. Sadly, by now I had already made up a new lead to correct the fault! The new firmware and the original lead now work extremely well together and, with the imaging and guiding under the control of MaximDL, I have a very workable system although there are still some minor 'differential flexure' issues to be resolved through the use of a separate guide and main telescope.